Electronic equipment often require extra cooling to transfer and dissipate the heat generated by the various components such as microprocessors, and the most commonly used mechanism for removing heat from a product such as a computer or server is via a motor-driven fan. In a single-motor fan assembly, the motor is a single point of failure. Often, when a fan fails, the computer must be shut down so that overheating does not occur, which would further damage high heat-producing components like the microprocessor. Alternatively, the computer may continue operating, but at a reduced capacity (e.g., reducing the processor speed to prevent overheating). Most fan failures are caused by motor failure.
Computers designed for high availability service, such as servers, add extra fans to compensate for the possibility of a fan failure. This prior art cooling system design paradigm increases the overall server cost in several ways: increased cost for a complete additional fan or fans, increased use of scarce real estate in the packaging with consequent limitations on design and layout options, and increased demand for power management subsystems. The need for additional space for the extra fan(s) will affect the thermodynamic cooling process, since the airflow will be different when driven from various locations in the packaging. When the fan system is configured so that two or more fans are in line axially, a further degradation of cooling effectiveness occurs because of the reduced airflow caused by the blockage of a failing or non-operating fan being in the way of the operating fan. In some cases, two fans may be operative at the same time, thus requiring synchronization systems. Thus the increased availability from prior art fan systems comes with various other costs, additional design burdens, or impairments to the overall product design.